RefME

» see refme.com as it lived, on the wayback machine

~ the goodbye ~

This arrived about a decade too late to be mine. I did my undergraduate and graduate work between roughly 2003 and 2010, and references in that period meant typing them out by hand at two in the morning from a style guide photocopy, checking the comma placement, then finding you had an extra source and renumbering everything. By the time an app existed that could do it from a photograph I was an engineer writing papers occasionally rather than a student writing them weekly, and the deadline panic the product was built around was already behind me.

RefME was incorporated in London on 13 February 2014 as ReferenceME Limited and the apps were publicly downloadable by late March. Tom Hatton was the founder and chief executive. The feature that made people pay attention was the scanner: you pointed your phone's camera at the barcode on the back of a library book, and the app resolved the ISBN and produced a finished citation in whichever of its reference styles your department demanded. The style count kept climbing in the marketing, from six thousand in 2014 to seven and a half thousand by the time it died.

Growth was fast in the way edtech growth was fast in 2015. The company raised a million in angel money in January 2014 and announced a five million dollar seed round on 8 April 2015, four million of it from Sunny Varkey's GEMS Education. It had 800,000 registered users that month and was adding ten thousand a day. By May 2016 it claimed more than a million users across 14,000 institutions, had shipped a Word add-in and a Chrome clipper, and had launched a paid tier. The Series A was planned and never happened.

Chegg bought the assets in December 2016. The 10-K line is short: certain assets of RefME Ltd., a privately held online learning company based in London, England, total fair value of the purchase consideration 1.8 million dollars, of which 800,000 was deferred cash paid out over four quarters. A company that had raised five million sold for roughly a third of it. Users found out by email on 26 January 2017, and the app stopped being RefME on 7 March.

~ the chronicle ~

~ what Chegg was actually buying ~

Seven months before the RefME deal, in May 2016, Chegg had paid 42.3 million dollars for Imagine Easy Solutions, which owned Cite This For Me and EasyBib. So by December it already had the citation generator it wanted. RefME was the competitor, and 1.8 million was the price of not having one. Everything that made RefME different went in the bin at migration: the Shibboleth institutional login, the Edmodo integration, the .RIS batch import and the folder structure all failed to make the trip, and university libraries wrote the loss up for their students in careful, disappointed blog posts.

~ the second deadline ~

The shutdown had a sting on a delay. Migrated accounts kept their saved references on Cite This For Me until 1 June 2017, and after that date saving anything required a subscription at £6.99 a month. So the free tool a million students had been using became a paid tool three months after the brand disappeared, which is the part of an acquisition that never makes the press release.

The announced date slipped once, too. The January email and the first wave of library notices all said 28 February 2017. By 20 February the University of Brighton was telling students it would be 7 March instead, and the archive confirms the cutover landed inside that day: the app host was still serving RefME's own files on the evening of 6 March and was redirecting to citethisforme.com by the small hours of the 8th.

~ the citation graph ~

Hatton was open about the business model in 2015, and it was not subscriptions. The value was in watching what everybody cited: a gold mine for publishers, he told EdSurge, to see what the end users are using in their content. A million students generating references is a live map of which papers and which books are actually being read across 14,000 institutions, updated hourly, which no publisher has and every publisher wants. That dataset is the thing a 1.8 million dollar asset purchase quietly includes, and nobody has ever said what became of it.

@}-->--

              )
             (_)
            .-'-.
            |   |
            |   |
            |   |
            |   |
          __|   |__   .-.
       .-'  |   |  `-:   :
      :     `---'     :-'
       `-._       _.-'
           '""""""

« previous  ·  random  ·  next »